DEVELOPING AI ASSISTANCE FOR INCLUSIVE COMMUNICATION IN ITALIAN FORMAL WRITING

Abstract

The push for inclusive language has gained global momentum, recognizing its crucial role in fostering respectful and equitable communication. While English has seen significant advancements in AI-assisted inclusive writing, highly gendered languages like Italian present unique challenges due to their complex grammatical structures, where gender is intrinsically linked to nouns, articles, adjectives, and verb participles. This article proposes a novel approach for developing an AI-powered system designed to assist in writing inclusive language within Italian formal communications. Our methodology integrates sophisticated Natural Language Processing (NLP) techniques, leveraging advanced Italian Language Models fine-tuned on expert-annotated datasets of formal texts. We outline the process of identifying non-inclusive patterns, generating grammatically correct and contextually appropriate inclusive alternatives, and the critical role of Explainable AI (XAI) to ensure user trust and understanding. Anticipated outcomes suggest that this system will significantly enhance the efficiency and accuracy of inclusive writing in Italian formal settings, promoting broader adoption of gender-neutral and person-first language principles.